Increased expression of HDJ-2 (hsp40) in carotid artery atherosclerosis: a novel heat shock protein associated with luminal stenosis and plaque ulceration.

PURPOSE: Evidence suggests that both humoral and cellular autoimmune processes directed toward heat shock proteins (hsp) contribute to the pathogenesis of atherosclerosis. We characterized a human hsp distinct from those previously characterized in atherosclerotic lesions, termed HDJ-2. METHODS: To determine the role of HDJ-2 in atherosclerosis, we compared the level of HDJ-2 mRNA expression with the level of hsp60 and hsp70 mRNA expression in 26 carotid endarterectomy specimens and 17 normal arteries. The level of expression of HDJ-2 mRNA was also correlated to the presence of plaque ulceration and the degree of luminal stenosis associated with the lesion. RESULTS: The expression of HDJ-2 and hsp70 was significantly higher in carotid artery plaques as compared with normal arteries: HDJ-2, 6.7 +/- 1.6 vs 0.1 +/- 0.04, (P =.001); hsp70, 9.5 +/- 0.9 vs 3.7 +/- 0.8, (P =.002). There was no significant difference in hsp60 expression between carotid artery plaques and normal arteries (21.0 +/- 0.9 vs 20.6 +/- 0.8, P =.65). Increased HDJ-2 expression in carotid artery plaques was independent of hsp70 (Pearson correlation, r = 0.11; Bartlett chi(2) analysis, P =.71). Within the ulcerated plaque group, there was a correlation between degree of stenosis and high HDJ-2 mRNA expression (r = 0.896, P =.016). However, there was no correlation between degree of stenosis and high HDJ-2 mRNA expression within the nonulcerated plaque group (r = 0.530, P =.076) or within the entire group of patients (r = 0.0085, P =.97). CONCLUSION: These results demonstrate that expression of HDJ-2 is significantly increased in atherosclerotic carotid artery plaques as compared with hsp60 and hsp70 and correlates with luminal stenosis in ulcerated atherosclerotic carotid artery plaques.

Preoperative treatment with doxycycline reduces aortic wall expression and activation of matrix metalloproteinases in patients with abdominal aortic aneurysms.

PURPOSE: Matrix metalloproteinases (MMPs) are considered to play a central role in the pathogenesis of abdominal aortic aneurysms (AAAs). Doxycycline (Dox) has direct MMP-inhibiting properties in vitro, and it effectively suppresses the development of elastase-induced AAAs in rodents. The purpose of this study was to determine if treatment with Dox suppresses MMPs within human aneurysm tissue and to elucidate the molecular mechanisms underlying this effect. METHODS: Aneurysm tissues were obtained from 15 patients with an AAA, eight of whom had been treated with Dox before surgery (100 mg orally twice a day for 7 days). Protein extracts were examined by means of gelatin zymography and immunoblot analysis, and RNA was examined by means of reverse transcription-polymerase chain reaction (RT-PCR). The effects of Dox on MMP production were further examined in human THP-1 mononuclear phagocytes in vitro. RESULTS: No detectable difference was found between groups by using substrate zymography as a means of assessing total MMP activity, but Dox treatment was associated with a slight (24.4%) reduction in the activated fraction of 72-kDa gelatinase (MMP-2; P <.05). In contrast, a 2.5-fold reduction in the amount of extractable 92-kDa gelatinase (MMP-9) protein in Dox-treated patients was revealed by means of immunoblot analysis (P <.05). Also, a 5.5-fold (81.9%) reduction in MMP-9 messenger RNA (mRNA) in Dox-treated patients was demonstrated by means of quantitative competitive RT-PCR (mean +/- SE, mol MMP-9/mol beta-actin: 1.3 +/- 0.5 vs 7.2 +/- 3.1; P <.04). There was no significant difference between groups in the relative expression of MMP-2 protein or mRNA. In cultured THP-1 monocytes stimulated with phorbol ester, the expression of MMP-9 protein and mRNA were both decreased after exposure to relevant concentrations of Dox in vitro. CONCLUSION: In addition to its recognized effects as a direct MMP antagonist, Dox may influence connective tissue degradation within human aneurysm tissue by reducing monocyte/macrophage expression of MMP-9 mRNA and by suppressing the post-translational processing (activation) of proMMP-2. Through this complementary combination of mechanisms, treatment with Dox may be a particularly effective strategy for achieving MMP inhibition in patients with an AAA.

Tc-99m sulfur colloid scintigraphy for detecting perigraft flow following endovascular aortic aneurysm repair: A feasibility study.

PURPOSE: To determine if scintigraphy with Tc-99m sulfur colloid can be used to detect perigraft flow after stent-graft repair of abdominal aortic aneurysm (AAA). METHODS: Twenty-three men and two women aged 56-84 years (mean 71 years) underwent endoluminal AAA repair as part of the EVT Phase II trial [EVT = Endovascular Technologies (Menlo Park, CA, USA)]. Aneurysm size averaged 5.4 cm (range 3-8 cm). Sixteen bifurcated, seven tube, and two aorto-uniiliac grafts were placed. Two days after stent-graft placement, patients underwent both contrast-enhanced computed tomography (CT), including delayed views, and Tc-99m sulfur colloid scintigraphy. RESULTS: Perigraft flow was found in only one patient at completion of angiography. Four additional patients had perigraft flow, discovered during their postoperative follow-up CT. Four patients had leaks at an attachment site and one had retrograde branch flow. Tc-99m sulfur colloid scintigraphy failed to diagnose any of the five leaks prospectively. In two of these patients, however, some abnormal paraaortic activity was noted in retrospect. CONCLUSION: Tc-99m sulfur colloid scintigraphy was unable to demonstrate endoleak with either rapid flow (attachment site leak) or slow filling (branch flow).

Source of elastin-degrading enzymes in mycotic aortic aneurysms: bacteria or host inflammatory response?

Elastolytic matrix metalloproteinases play a central role in the development of chronic atherosclerotic aortic aneurysms, but mycotic aortic aneurysms are a distinct and unusual form of aneurysm disease caused by bacterial infection. Mycotic aortic aneurysms follow a more rapid and unpredictable course than chronic aneurysm disease and they exhibit a predilection for the suprarenal aorta, further implying unique pathophysiologic mechanisms. The purpose of this study was to examine the nature and source of elastin-degrading enzymes in mycotic aortic aneurysm. Bacterial isolates and aortic tissues were obtained from four consecutive patients undergoing surgical repair of suprarenal mycotic aortic aneurysm. Using an in vitro 3H-labeled elastin degradation assay, elastin-degrading enzyme activity was only observed in the bacteria-conditioned medium from an isolate of Pseudomonas aeruginosa. Elastin-degrading enzyme activity in the aortic tissue homogenate of this patient was abolished by the serine protease inhibitor, phenylmethylsulfonyl fluoride, but it was not suppressed by the metalloproteinase inhibitor, ethylenediamine tetraacetic acid (EDTA). In contrast, elastin-degrading enzyme activity in the bacterial-conditioned medium was decreased by about half by both phenylmethylsulfonyl fluoride and EDTA. Elastin substrate zymography revealed two phenylmethylsulfonyl fluoride-inhibitable elastin-degrading enzyme activities in the aortic tissue homogenate that corresponded to human neutrophil elastase (approximately 30 kDa) and its stable complex with alpha 1-proteinase inhibitor (approximately 80 kDa), but no activity attributable to Pseudomonas elastase, a 33-kDa metal-dependent enzyme. Human neutrophil elastase was readily detected throughout mycotic aortic aneurysm tissues by immunohistochemistry, but elastolytic metalloproteinases were only occasionally observed. The results of this study suggest that the elastin-degrading enzyme produced in mycotic aortic aneurysm are largely serine proteases of host neutrophil origin, rather than elastases produced by the infecting microorganisms or the macrophage-derived metalloproteinases typically observed in atherosclerotic aneurysm disease. Further studies will be needed to extend these findings to a larger number of patients with mycotic aortic aneurysm and those caused by additional microorganisms.

Recurrent gastrointestinal bleeding: use of thrombolysis with anticoagulation in diagnosis.

PURPOSE: To determine the safety and diagnostic accuracy of a provocative protocol with heparin and urokinase to induce bleeding and determine the source in patients with chronic gastrointestinal hemorrhage. MATERIALS AND METHODS: Nine patients had gastrointestinal bleeding from an indeterminate source and had negative results from esophagogastroduodenoscopy, colonoscopy, small-bowel examination, and angiography. Ten provocative bleeding studies were performed prospectively. Patients had no clinical evidence of bleeding within 2 days before the study. Intravenous administration of heparin and urokinase was performed systemically during a 4-hour period while scintigraphy was performed continuously. Mesenteric angiography was performed immediately in patients in whom substantial gastrointestinal bleeding was detected at scintigraphy. RESULTS: The provocative protocol was successful in inducing scintigraphically detectable hemorrhage in four (40%) studies within 4 hours. In two of these four studies, the source of hemorrhage was determined and treated with embolization or surgery. Three (30%) studies demonstrated scintigraphic evidence of hemorrhage only at delayed imaging (8-24 hours after initiation of the study). The remaining three (30%) studies did not show active bleeding. No complications occurred, including hemodynamic instability or uncontrollable decreases in hematocrit. CONCLUSION: Since this protocol with heparin and urokinase enabled determination of the bleeding source in only two of 10 studies, protocol modifications are necessary before this intervention is used widely.

Surgical treatment of patients with symptomatic vertebrobasilar insufficiency.

We report our experience with surgical management of symptomatic vertebrobasilar insufficiency (VBI). Forty revascularizations were carried out in 39 patients over 90 months. Dizziness (52%) and syncope/presyncope (32%) were the most common symptoms. Arteriography was performed in all patients, with subclavian steal seen in 55% of patients. Procedures performed included 22 cases of carotid-subclavian bypass or transposition (55%), seven direct vertebral reconstructions (17.5%), four great vessel reconstructions (10%), four isolated carotid endarterectomies (10%), and three axilloaxillary bypasses (7.5%). One patient died, and the combined morbidity and mortality rate was 15%. Outpatient follow-up was available on 37 of the 38 patients discharged alive. At a mean follow-up of 16.4 months, 34 patients had no VBI complaints. Three of four patients treated with CEA alone had persistent VBI complaints. We conclude that a variety of anatomic lesions can result in VBI symptoms, with subclavian steal being the most common. Procedures which directly correct the anatomic abnormality result in sustained symptom resolution with acceptable complication rates.

Endovascular stent grafts for aneurysmal and occlusive vascular disease.

BACKGROUND: This report details our initial experience with two types of endovascular grafts- one for the treatment of infrarenal abdominal aortic aneurysms and the other for the treatment of iliac artery occlusive disease. METHODS: An abdominal aortic aneurysm was repaired in 34 patients using 3 different types of Ancure endografts (Menlo Park, California). Control patients (n = 9) had a standard aneurysm repair. Patients with chronic lower extremity ischemia (n = 7) secondary to iliac artery occlusive disease were treated with a Hemobahn endograft (W. L. Gore & Associates, Flagstaff, Arizona). RESULTS: Ancure graft deployment was achieved in 33 of 34 (97.1%) patients. Perioperative mortality for the Ancure and control group patients was 2.9% and 0%, respectively. Periprosthetic leaks were identified within 48 hours of deployment in 6 (18.2%) Ancure graft patients. All but 2 of the leaks resolved on serial follow-up. Additional endovascular procedures were required in 11 (32.4%) Ancure graft patients at the initial procedure or during follow-up to correct graft or arterial stenoses. Patients treated with an endovascular graft had significantly less blood loss and shorter hospital stays than control group patients. For Hemobahn patients, the technical success for graft deployment was 100%. There were no perioperative deaths. The ankle/brachial index increased from a mean of 0.52 preoperatively to 0.86 postoperatively (P = 0.004). One patient required a Wallstent in follow-up to correct a graft wrinkle. Angiography at 6 months demonstrated mild intimal hyperplasia in the stent graft in 5 of 6 patients. CONCLUSIONS: These early results support the potential for endovascular grafts in the treatment of aneurysmal and occlusive vascular disease. Further modifications in the devices and deployment techniques are necessary to reduce the incidence of periprosthetic leaks, graft limb stenoses, and intimal hyperplasia.

Transient obstruction of the internal carotid artery during angiography.

While performing carotid angiography in a 76-year-old man, we found that the right internal carotid artery repeatedly opened and closed during the examination. The patient experienced no related neurologic events. The explanation, confirmed at surgery, was that a flap associated with an atherosclerotic plaque had acted as a ball valve.

Spontaneous closure of selected iatrogenic pseudoaneurysms and arteriovenous fistulae.

PURPOSE: We report our approach to the management of postcatheterization femoral artery pseudoaneurysms and arteriovenous fistulae in an attempt to determine the frequency of spontaneous resolution of selected lesions. METHODS: We studied 196 pseudoaneurysms, 81 arteriovenous fistulae, and 9 combined lesions that were identified by duplex scan. Indications for immediate surgical repair included pseudoaneurysm greater than 3 cm, enlarging hematoma, pain, groin infection, nerve compression, limb ischemia, concomitant surgical procedure, and patient refusal or inability to comply with follow-up. All other lesions were observed. RESULTS: One hundred thirty-nine patients underwent prompt surgical repair, and 147 patients were initially managed without operation. There were no limb-threatening complications associated with nonoperative management in this subset of patients. Eighty-six percent of the lesions being observed resolved spontaneously within a mean of 23 days, whereas 14% required surgical closure for a variety of reasons (at a mean of 111 days after the initial diagnosis). There was no statistically significant difference in the rate of spontaneous pseudoaneurysm closure (89%) as opposed to fistulae (81%) (p < 0.17). By life-table analysis, 90% of selected pseudoaneurysms had resolved by 2 months. Patients selected for observation underwent an average of 2.6 duplex scans per patient versus 1.4 scans per patient for those treated with immediate surgery (p < 0.01). CONCLUSION: The natural history of stable pseudoaneurysms and arteriovenous fistulae is benign and frequently results in spontaneous resolution, which allows properly selected patients to be managed without operation.

Outcome in medical intensive care unit patients requiring abdominal surgery: prospective validation of a risk classification system.

To prospectively validate a previously developed two-factor logistic regression model as a predictor of mortality, we studied its effectiveness in predicting outcome for patients in medical intensive care units (ICUs) who had surgical laparotomy. A high-risk classification was assigned to patients with an Organ System Failure Index (OSFI) > or = 3 or an APACHE (Acute Physiology and Chronic Health Evaluation) II score > 18 within 24 hours of surgery. The in-hospital mortality rate of surgical patients classified as high risk (n = 32) was significantly greater than that of surgical patients classified as low risk (n = 42) (62.5% versus 9.5%; relative risk, 6.6; 95% confidence interval, 2.5 to 17.3). Mortality after surgery correlated with presence or absence of the two variables from the logistic regression model: neither present, 9.5%; APACHE II > 18 present, 68.0%; OSFI > or = 3 present, 75.0%; both present, 88.2%. We showed that a two-factor risk classification at the time of surgical evaluation can be used to stratify medical ICU patients according to risk of in-hospital mortality.

Therapeutic footwear can reduce plantar pressures in patients with diabetes and transmetatarsal amputation.

OBJECTIVE: To compare how footwear (full-length shoe or short shoe), a total contact insert, a rigid rocker-bottom (RRB) sole, and an ankle-foot orthosis (AFO) affect peak plantar pressure (PPP) on the distal residuum and contralateral extremity of patients with diabetes and transmetatarsal amputation (TMA). RESEARCH DESIGN AND METHODS: Thirty patients with diabetes and TMA participated (mean age 62 +/- 4 years). In-shoe plantar pressures during walking were measured in six types of footwear. Each measurement occurred after a 1-month adjustment period. Repeated measure analysis of variance (ANOVA) was used to compare treatments. RESULTS: All five types of therapeutic footwear reduced plantar pressures compared with regular shoes with a toe-filler (P < 0.05). A full-length shoe, total contact insert, and RRB sole resulted in lower pressures on the distal residuum (222 vs. 284 kPa) and forefoot of the contralateral extremity (197 vs. 239 kPa), compared with a regular shoe and toe-filler. Footwear with an AFO showed reduced PPP on the residuum, but most patients complained of reduced ankle motion during walking. A short shoe reduced pressures on the residuum, but not on the contralateral extremity, and many patients had complaints regarding cosmesis of the shoe. CONCLUSIONS: The full-length shoe, total contact insert, and an RRB sole provided the best pressure reduction for the residuum and contralateral foot, with the optimal compromise for cosmetic acceptance and function.

Revascularization of the solitary kidney: a challenging problem in a high risk population.

BACKGROUND: Patients with significant atherosclerotic stenosis involving the artery to a solitary functioning kidney present a clinical challenge. METHODS: From August 1987 through August 1995, 35 of these patients (average age, 68.4 +/- 6.9 years) were treated. Comorbid conditions included previous myocardial infarction in 23% of the patients, congestive heart failure (CHF) in 34%, chronic obstructive pulmonary disease in 20%, and diabetes in 20%. The average creatinine level of the patients was 2.5 +/- 1.5 mg/dl. Indications for revascularization were hypertension in 86%, hypertensive crisis with CHF in 17%, and renal insufficiency in 69%. Procedures performed included 19 extra-anatomic bypasses, 8 concomitant with infrarenal aortic reconstruction and 2 concomitant with thoracoabdominal aortic aneurysm repair; 1 visceral segment endarterectomy; 1 renal artery endarterectomy with reimplantation; I superior mesenteric to renal artery bypass; 1 aortorenal bypass; and 2 percutaneous angioplasties with staged nephrectomies. RESULTS: At discharge, 91% of patients had stable or improved renal function with an average creatinine level of 1.7 +/- 0.8 mg/dl. Hypertension was cured or improved in 85%. Perioperative mortality was 6%, and major morbidity was 43%, including the need for permanent (9%) and temporary (9%) dialysis, respiratory insufficiency (18%), two early reoperations, six cardiac complications, one case of gastrointestinal bleeding, and one stroke. In the follow-up period (mean duration, 39.2 months), survival has been 73%, and no additional patients have required dialysis. CONCLUSIONS: Although significant perioperative morbidity exists in this high risk population, the long-term preservation of renal function and improvement in hypertension make solitary renal revascularization worthwhile.

Autologous vein-covered stent for the endovascular management of an iliac artery-ureteral fistula: case report and review of the literature.

Iliac artery-ureteral fistula is a rare entity that is being reported with increasing frequency. Patients with iliac artery-ureteral fistulas can be divided into two distinct groups on the basis of the factors that predispose them to having these fistulas. In group I the fistula is associated with degenerative iliac artery disease or previous arterial reconstructive surgery. Patients in group II have undergone some combination of the following procedures: pelvic extirpative surgery for malignancy, urinary diversion, radiation therapy, and ureteral stenting. The diagnosis of an iliac artery-ureteral fistula can be elusive even with the use of multiple imaging methods. Direct operative repair is technically demanding and is associated with high mortality rates. In recent years, treatment has shifted toward percutaneous embolization of the iliac artery and extraanatomic lower extremity vascular reconstruction for group II patients. In this report, the 24 group II patients with iliac artery-ureteral fistulas who previously have been described are reviewed, and a new endovascular treatment for this entity that uses a stented vein graft is detailed.

Femoropopliteal bypass for claudication: vein vs. PTFE.

The vascular graft of choice for femoropopliteal bypass in patients with intolerable claudication is controversial. We retrospectively reviewed our experience with 239 patients suffering from claudication secondary to superficial femoral artery obstruction. Femoropopliteal reconstruction was performed with saphenous vein to the below-knee popliteal artery in 66 patients (BK-vein). Polytetrafluoroethylene (PTFE) was used in 128 patients as a bypass graft to the above-knee popliteal artery (AK-PTFE) and 45 patients had a PTFE graft to the below-knee popliteal artery (BK-PTFE). All patients were enrolled in a postoperative graft surveillance program with graft revision when appropriate. There was one perioperative death (0.4%). Primary patency at 5 years for AK-PTFE, BK-PTFE, and BK-vein was 58.0%, and 60.3%, respectively, and was not significantly different among the graft groups. Graft revision for failed/failing grafts resulted in 5-year secondary patency rates of 79.2% (AK-PTFE), 73.3% (BK-PTFE), and 74.4% (BK-vein). These secondary patency rates were not statistically different. Eventual conversion to a vein graft in patients initially treated with PTFE maximized patency in the femoropopliteal segment with 5-year patency rates of 84.6% and 93.0% for the AK-PTFE and BK-PTFE graft groups, respectively. Major leg amputation was necessary during the entire course of the study in eight (3.3%) patients. We conclude that long-term patency rates for femoropopliteal bypass in patients with intolerable claudication are similar for PTFE and autologous saphenous vein grafts.

Role of Doppler US in screening for carotid atherosclerotic disease.

PURPOSE: To evaluate Doppler ultrasound (US) as a screening modality before arteriography for extracranial carotid artery disease. MATERIALS AND METHODS: The net benefit in stroke reduction from screening with Doppler US was calculated on the basis of literature estimates of disease prevalence, risk reduction data, and locally validated sensitivities and specificities for detection of carotid artery stenosis and occlusion in 215 patients. RESULTS: Screening a symptomatic population demonstrated a net stroke reduction. Screening asymptomatic populations with a 20% prevalence of > or = 60% stenosis also yielded a net stroke reduction. Screening low-prevalence (5%) asymptomatic populations produced a small benefit, which was lost if arteriographic or surgical complications increased slightly. Arteriographic confirmation of the US diagnosis of occlusion produced a small benefit only in the symptomatic population. CONCLUSION: Screening symptomatic and high-prevalence asymptomatic populations with US reduces stroke. Increased arteriographic or surgical complication rates reduce the benefit of screening in any population.

Limb salvage and patency after aortic reconstruction in younger patients.

PURPOSE: The purpose of this report was to compare patient characteristics, treatment, and outcome in younger and older patients with aortoiliac occlusive disease. METHODS: The medical records of 56 patients < or = 50 years of age (Group < or = 50) were retrospectively reviewed and compared to the records of 128 patients > or = 60 years of age (Group > or = 60). All patients were examined and treated between April 1987 and April 1994. Postoperatively, they were enrolled in a vascular laboratory surveillance program to serially monitor the status of the vascular reconstruction. Follow-up averaged more than 3 years in both groups and was available on greater than 90% of patients. RESULTS: Patients in Group < or = 50 had a higher incidence of smoking (68% versus 51%) and a lower incidence of hypertension (29% versus 50%) than patients in group > or = 60 (smoking P = 0.03, hypertension P = 0.007). No other significant differences were noted among cardiovascular risk factors. Preoperative indications for surgery were similar among patients in both groups. An aortoiliac endarterectomy was more commonly used to revascularize the lower extremities in younger patients than in their older counterparts (23% versus 7%, P = 0.002). Graft revisions were more frequently necessary after aortic reconstruction in Group < or = 50 than in Group > or = 60 (29% versus 8%, P = 0.0003); however patency rates computed by life table analysis were not significantly different. The primary patencies for Group < or = 50 and Group > or = 60 at 5 years were 64% and 67%, respectively; their secondary patency rates at 5 years were 84% and 89%, respectively. No significant difference was found in major limb amputation (8% in Group < or = 50 versus 5% in Group > or = 60, P = 0.46). We conclude that aortoiliac reconstruction for occlusive disease can be performed with similar secondary patency and amputation rates in young and old patients. However, close postoperative surveillance and frequent surgical revision are necessary to maintain patency and minimize amputation.

Relationship between evoked potentials and clinical status in spinal cord ischemia.

STUDY DESIGN AND METHODS: Sciatic neurogenic motor-evoked potentials, spinal evoked potentials, and somatosensory-evoked potentials were recorded in 12 anesthetized dogs that had arterial ischemia of the lumbar cord produced by ligation of segmental arteries. The presence or absence of the above-mentioned potentials was compared with the clinical status of repeated wake-up tests. RESULTS: Although these results were complicated, sciatic neurogenic motor-evoked potential was more sensitive to the spinal cord ischemia and was a better predictor of clinical outcome than spinal evoked potential and somatosensory-evoked potential. However, the presence was not a guarantee of normal function. The initial morphologic change of these potentials secondary to ischemia consisted of a decrease in amplitude and in the number of peaks without a shift of latency. CONCLUSIONS: The present study suggests that the peripheral neurogenic motor-evoked potential is a better warning system for spinal cord ischemia and that its adoption may contribute to the prevention of cord ischemia during spinal surgery, whereas somatosensory-evoked potential and spinal evoked potential cannot be indices.

Transabdominal versus retroperitoneal incision for abdominal aortic surgery: report of a prospective randomized trial.

PURPOSE: The purpose of this study was to perform a randomized, prospective trial that compares the transabdominal with the retroperitoneal approach to the aorta for routine infrarenal aortic reconstruction. METHODS: From August 1990 through November 1993, patients undergoing surgery for abdominal aortic aneurysm (AAA) disease or aortoiliac occlusive disease (AIOD) were asked to participate in a randomized trial comparing the transabdominal incision (TAI) to the retroperitoneal incision (RPI) for aortic surgery. One hundred forty-five patients were randomized, with 75 (41 with AAA and 34 with AIOD) in the TAI group and 70 (40 with AAA and 30 with AIOD) in the RPI group. There were no significant differences between the groups in terms of age, sex, postoperative pain control (epidural vs patient-controlled analgesia), or comorbid conditions, except for a higher incidence of chronic obstructive pulmonary disease in the TAI group (21 vs 8 patients). RESULTS: The incidence of intraoperative complications was similar for both groups. After surgery, the incidence of prolonged ileus (p = 0.013) and small bowel obstruction (p = 0.05) was higher in the TAI group. Overall, the RPI group had significantly fewer complications (p < 0.0001). The overall postoperative mortality rate (two deaths) was 1.4%, with both occurring in the TAI group (p = 0.507). The RPI group also had significantly shorter stays in the intensive care unit (p = 0.006), a trend toward shorter hospitalization (p = 0.10), lower total hospital charges (p = 0.019), and lower total hospital costs (p = 0.017). There was no difference in pulmonary complications (p = 0.71). In long-term follow-up (mean 23 months), the RPI group reported more incisional pain (p = 0.056), but no difference was found in incisional hernias or bulges (p = 0.297). CONCLUSIONS: We conclude that the RPI approach for abdominal aortic surgery is associated with fewer postoperative complications, shorter stays in the hospital and intensive care unit, and lower cost. There is, however, an increase in long-term incisional pain. Current methods of postoperative pain control seem to decrease the incidence of pulmonary complications.